University of Groningen
Outcomes after Spinal Cord Injury
Osterthun, Rutger
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Osterthun, R. (2018). Outcomes after Spinal Cord Injury. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
4
Causes of death following spinal cord
injury during inpatient rehabilitation
and the first five years after
discharge. A Dutch cohort study
Rutger Osterthun, MD1,Marcel WM Post, PhD2, Floris WA van
Asbeck, MD, PhD3, Christel MC van Leeuwen, PhD2, Casper F van
Koppenhagen MD, PhD3
1Center of Excellence for Rehabilitation Medicine, University Medical Center Utrecht and De
Hoogstraat Rehabilitation, Utrecht, the Netherlands
2Brain Center Rudolf Magnus and Center of Excellence for Rehabilitation Medicine, University
Medical Center Utrecht and De Hoogstraat Rehabilitation, Utrecht, the Netherlands
3Department of Spinal Cord Management, De Hoogstraat Rehabilitation, Utrecht, the Netherlands
aBSTRaCT
design: Prospective multicentre cohort study.
Objectives: To determine mortality, causes and determinants of death of individuals with spinal cord injury (SCI) within five years after first inpatient rehabilitation.
Setting: The Netherlands.
methods: Patients were included on admission to first clinical rehabilitation after traumatic or non-traumatic SCI. Inclusion criteria: age between 18 and 65, ASIA impairment scale A-D, and expected long-term wheelchair dependency. Information about survival, cause of death, relevant co-morbidity and psychosocial circumstances was obtained from the rehabilitation physician or general practitioner. Determinants of death were retrieved from a prospectively collected database. Deceased persons and survivors were compared using Chi-square and t-tests. Cox regression analysis was performed to describe independent predictors of death. The Kaplan-Meier method was used to calculate survival curves for independent predictors. Excess mortality was described by a Standardized Mortality Ratio (SMR).
Results: Mean duration of follow-up was 6.2 years. A total of 27 persons (12.2%) deceased during this period (SMR 5.3). Main causes of death were cardiovascular dis-ease (37.0%), pulmonary disdis-ease (29.6%) and neoplasm (14.8%). Older age at injury, non-traumatic SCI, family history of cardiovascular disease, less social support and a history of other medical conditions on admission were related to death. Older age at injury, non-traumatic SCI and a history of other medical conditions were independent predictors of death.
Conclusion: Twelve per cent of persons with SCI who had survived the acute hospital phase died during follow-up (SMR 5.3). The main causes of death were cardiovascular and pulmonary disease.
INTROduCTION
Although survival of persons with spinal cord injury (SCI) has improved considerably in the last decades,1,6,7 there remains an increased risk of premature death due to, for example, pneumonia or suicide.1-5 The main long-term causes of death of persons with SCI have changed from urological complications, to cardiovascular and pulmonary dis-eases.1 Several studies on mortality after SCI have been performed.1 Comparing results of these studies is however difficult due methodological differences, such as case selec-tion and survival periods. Furthermore, most of these studies restricted their samples to persons with traumatic SCI and most focus on long-term mortality.1-5 It has been suggested that the decline in mortality since the 1980s is mainly based on improvement of acute care.6-7 The risk of death is however still the highest during the first year(s) after injury.5,7,8 In addition to personal and lesion characteristics, several psychosocial and life style factors have been found to be associated with death.2,9,10
In the Netherlands, persons with a new SCI are generally admitted to the hospital for a few days to weeks. When their medical condition has stabilized, they are transferred to a nursing home or a rehabilitation centre with a specialized SCI unit to continue their inpatient rehabilitation. Some information on mortality in the acute care hospital is available,11 but post-acute mortality and causes of death of persons with SCI have not previously been described in the Netherlands. Internationally, only limited information is available about the causes of death in the first years after SCI for persons who survived the hospital phase.3,4,12 Understanding the causes of death in this period can yield relevant information for further improvements of care.
The purpose of this study is to describe mortality, causes and determinants of death between the start of first inpatient SCI rehabilitation until five years after discharge in the Netherlands. We hypothesized that besides personal and lesion characteristics, health variables, life style and psychosocial factors are associated with mortality.
mEThOdS
Study population
The study population consisted of 225 individuals with traumatic and non-traumatic SCI included in the Dutch prospective multicentre cohort study “Restoration of (wheel-chair) mobility in spinal cord injury rehabilitation”.13 The observation period started on admission of their first inpatient rehabilitation in eight rehabilitation centres with specialised SCI units in the Netherlands between August 2000 and July 2003. Further inclusion criteria of the cohort were: age between 18 and 65 years, grade A, B, C or D on the American Spinal Injury Association (AISA) Impairment Scale (AIS) and expected
wheelchair dependency. Participants were excluded if their SCI was caused by a ma-lignant tumour, if they had a progressive disease, psychiatric problems or insufficient command of the Dutch language to participate in the study.
Procedure
Information was prospectively collected at the start of active rehabilitation, defined as the moment that the participant was able to sit in a wheelchair for 3-4 hours, by a medical examination, an oral interview by a trained research assistant and self-report questionnaires.
Additional information on survival status and, if applicable, cause of death was requested retrospectively from the rehabilitation physician with a form developed for this project. If this procedure yielded insufficient information, the participant’s general practitioner was approached.
The research protocol of the cohort study was approved by the Medical Ethics Committee of the SRL/iRv. All persons gave written informed consent. The additional collection of data on survival and causes of death was approved by the local medical ethics committee of De Hoogstraat Rehabilitation, Utrecht, the Netherlands.
Instruments
Survival status and causes of death: with a structured self-developed form, information was asked about the survival status, cause of death based on the ICD-10, date of death, relevant co-morbidity and psychosocial circumstances that had possibly contributed to death.
Lesion characteristics: level and completeness of SCI were assessed according to the International Standards for Neurological Classification of Spinal Cord Injury.14 Functional independence was measured with the motor score of the Functional Independence Measure (FIM), consisting of 13 items about self-care, mobility, transfers, and toileting. The FIM motor score is a responsive measure in persons with SCI.15
Psychosocial factors: social support was measured with the Social Support List, which is a reliable and valid short version of the Social Support List-Interactions, consisting of 12 items.16 A total score is calculated by adding up all items (range 12–48). Mental health and vitality were assessed using the Mental Health and Vitality subscales of the 36-Item Short Form Health Survey (SF-36), consisting of respectively 5 and 4 items. The SF-36 has shown acceptable reliability and validity in persons with SCI.17 Life satisfaction was measured with the LS questions, which have shown good validity in measuring life satisfaction in persons with SCI.18
Health factors: whether persons had a history of other medical conditions or a family history of cardiovascular disease was assessed by the research assistant. A history of other medical conditions included all other medical conditions before admission to
the rehabilitation centre, other injuries at lesion onset and in-hospital complications. Secondary SCI conditions were not included. Body mass was assessed by the research assistant and body height was self-reported. The Body Mass Index (BMI) was computed and being overweighed was defined as a BMI > 25.
Lifestyle factors: doing sports before SCI, smoking and drinking alcohol before SCI were self-reported on admission.
data analysis
Analyses were performed using IBM SPSS Statistics V.19. Differences between de-ceased persons and survivors were described using the Chi-square test and t-test. A cox regression analysis was performed to describe independent predictors of death. The Kaplan-Meier method was used to calculate the observed survival curves for indepen-dent predictors. Age at injury was dichotomized (age < 45 years and age ≥ 45 years) to enable the calculation and display of survival curves for this variable. This cut-off point was chosen to create more or less equal-sized groups.
The excess mortality was described by calculating a Standardized Mortality Ratio (SMR). This is the ratio of the observed patient mortality and the expected mortality in the Dutch population. The number of expected deaths during the time interval of the study, adjusted by sex and age, was calculated using the database of Statistics Netherlands (www.cbs.nl).
RESulTS
general characteristics
Survival status was successfully identified in 222 persons (98.7%). The remaining 3 persons were excluded. Mean time between injury and admission to the rehabilitation centre, i.e. time from injury to start of the observation period, was 45.8 (SD 43.4) days (median 32 days, range 0-281). Mean age at injury was 40.6 (SD 14.1) years (median 40.3 years, range 18.1-65.7), 74.3% were male, 59.9% had paraplegia, 69.1% had a motor complete lesion and 73.0% had a traumatic lesion. Accidents (69.2%) and falls (22.2%) were the most common causes of traumatic SCI. The causes of non-traumatic SCI were vascular disease (21.7%), inflammation (15.0%), iatrogenic (13.3%), benign tumour (10.0%) and other non-traumatic causes (40.0%).
A total of 27 persons (12.2%) died during the mean follow-up period of 6.2 (SD 1.3) years. With a number of expected deaths in the general population of 5.1, the SMR was 5.3 (95% CI 3.6-7.6). Mean age at death was 53.5 (SD 13.2) years. Mean time between injury and death was 4.1 (SD 2.7) years. The survival for persons with traumatic and non-traumatic SCI was 93.2% and 73.3% respectively. Mortality rates in
the different subgroups of non-traumatic SCI did not significantly differ, although there was a tendency towards higher mortality rates in subgroups with vascular and “other” non-traumatic causes of SCI. Further characteristics of deceased persons and survivors are displayed in table 1.
Table 1. Characteristics of deceased persons and survivors
Characteristics Deceased (%) (N=27) Survivors (%) (N=195) Total (%) (N=222) Chi square p-value Sex Male 74.1 74.4 74.3 0.001 0.975 Female 25.9 25.6 25.7 Marietal status Together 29.6 24.0 24.7 0.410 0.522 Alone 70.4 76.0 75.3 Education Low 40.7 32.5 33.5 9.078 0.215 Middle 33.3 50.8 48.6 High 25.9 16.8 17.9 Cause of injury Traumatic 40.7 77.4 73.0 16.192 0.000 Non-traumatic 59.3 22.6 27.0 Level of injury Tetraplegia 44.4 39.5 40.1 0.243 0.622 Paraplegia 55.6 60.5 59.9 Completeness of injury Motor Complete 72.0 68.7 69.1 0.112 0.738 Motor Incomplete 28.0 31.3 30.9 Neurological classification Incomplete paraplegia 16.0 16.9 16.8 0.569 0.904 Complete paraplegia 40.0 43.6 43.2 Incomplete tetraplegia 12.0 14.4 14.1 Complete tetraplegia 32.0 25.1 25.9
Family history of cardiovascular disease
Yes 46.2 24.5 27.1 5.450 0.020
No 53.8 75.5 72.9
History of other medical conditions
Yes 77.8 44.1 48.2 10.772 0.001
Causes of death
The characteristics of deceased persons are displayed in table 2. The main causes of death were cardiovascular disease (37%), pulmonary disease (29.6%) and neoplasm (14.8%). Three persons (23, 24 and 39 years of age, all with motor complete paraplegia) died during the first year after injury, all from pulmonary embolism, of whom two during inpatient rehabilitation. For one person, the heparin anticoagulation had been restarted a few days before death. The other two persons had used oral anti-coagulation during their admission, of which one until death. Seven persons died due to other cardiovascular disease of which six had a non-traumatic lesion. Of the eight persons who died due to pulmonary disease, six had a traumatic SCI and five had a tetraplegia. One death occurred due to homicide. This person had survived a previous assault that had resulted in the SCI. Psychosocial circumstances had possibly contributed to death in 22% of cases.
Table 1. Characteristics of deceased persons and survivors (continued)
Characteristics Deceased (%) (N=27) Survivors (%) (N=195) Total (%) (N=222) Chi square p-value
Being overweighed (BMI > 25)
Yes 26.1 28.7 28.4 0.070 0.791
No 73.9 71.3 71.6
Doing sports before injury
Yes 51.9 65.1 63.5 1.793 0.181
No 48.1 34.9 36.5
Drinking alcohol before injury
Yes 81.5 76.0 76.7 0.392 0.531
No 18.5 24.0 23.3
Smoking before injury
Yes 46.2 42.7 43.1 0.111 0.739
No 53.8 57.3 56.9
Mean (SD) Mean (SD) Mean (SD) t-test p value
Age at injury (years) 50.1 (11.9) 39.3 (13.9) 40.6 (14.1) 3.849 0.000
Social support (range 12-48) 32.0 (5.5) 35.6 (5.9) 35.1 (6.0) -2.833 0.005
Mental health (range 0-100) 63.7 (20.9) 66.9 (18.0) 66.5 (18.3) -0.823 0.412
Vitality (range 0-100) 56.8 (25.5) 58.3 (17.0) 58.1 (18.1) -0.286 0.777
FIM Motor score (range 13-91) 38.9 (18.3) 41.2 (19.0) 40.9 (18.9) -0.569 0.570
Table 2. Characteristics of deceased persons (N = 27)
Characteristic Traumatic (N) Non-traumatic (N) Total (N) Total (%)
Cause of death
Cardiovascular disease
Pulmonary embolism 2 1 3 11.1
Other cardiovascular disease 1 6 7 25.9
Airway/pulmonary
Sepsis airway 2 0 2 7.4
Influenza and pneumonia 2 2 4 14.8
Other pulmonary disease 2 0 2 7.4
Neoplasm 0 4 4 14.8
Peritonitis 0 1 1 3.7
Sepsis due to pressure sore 1 0 1 3.7
Euthanasia 0 1 1 3.7
Homicide 1 0 1 3.7
Unknown 0 1 1 3.7
Living situation at moment of death
Independent with partner 13 48.1
Independent without partner 4 14.8
Independent living with regular help 2 7.4
Nursing home 1 3.7 Other 5 18.5 Unknown/missing 2 7.4 Location of death Home 12 44.4 Hospital 8 29.6 Rehabilitation centre 2 7.4 Nursing home 2 7.4 Other 1 3.7 Unknown/missing 2 7.4 Relevant co-morbidity Yes 20 74.1 No 4 14.8 Unknown/missing 3 11.1
Psychosocial circumstances possibly contributing to death
Yes 6 22.2
No 16 59.3
determinants of death
Older age at injury, non-traumatic cause of injury, family history of cardiovascular disease, less social support and having a history of other medical conditions were related to death (table 1). Lesion characteristics were not related to death, neither when calculated separately for traumatic and non-traumatic SCI.
In cox regression analyses, age at injury, non-traumatic etiology and having a history of other medical conditions were independent predictors of death (table 3). Calculated survival curves for independent predictors are shown in figure 1-3.
Table 3. Cox regression analyses of predictors of death
Predictor Relative Risk 95% CI p-value
Cause of injury (non-traumatic) 4.0 1.6-10.4 0.004
Age at injury (older) 1.0 1.0-1.1 0.046
Having a history of other medical conditions (yes) 5.7 1.9-16.7 0.001
figure 2. Survival curves for history of other medical conditions
dISCuSSION
A high number of persons with SCI who had survived the acute hospital phase died during the 6.2 year follow-up (12%, SMR 5.3), with cardiovascular and pulmonary diseases as major causes of death (65%). Age at injury, non-traumatic etiology and a history of other medical conditions were independent predictors of death.
Comparing the results of our study with other studies is difficult due to methodological differences. As all participants in our study had survived the acute hospital phase, our results are applicable to the rehabilitation phase and the first years afterwards. Previously we described an in-hospital mortality rate of 14% in traumatic SCI.11 Considering the generalization of our results, the participants of this study were on average younger, had more often a complete and traumatic SCI than persons with SCI admitted to Dutch and Flemish rehabilitation centres.19 The inclusion criteria of our study thus render a positive selection and, consequently, probably an underestimation of mortality compared to the Dutch SCI population.
Hartkopp et al. described mortality and causes of death of a traumatic SCI population that had also survived the acute phase (days to several months). Our SMR was more than twice as high as in their study, probably due to the inclusion of non-traumatic SCI in our study. Hartkopp et al. described a 10-year survival rate of 86.8%,4 which seems to correlate with our 93.2% in the subgroup of persons with traumatic SCI after a mean follow up of 6.2 years. Mesard et al. studied mortality of persons with traumatic SCI between 1955 and 1965. This study found a high mortality rate in the first 3 months and a 5-year survival of about 91% for persons who had survived the first 3 months.20 An Israeli study in a large group of persons with non-traumatic SCI described 5-year and 10-year survival rates of respectively 93.7% and 84.2%.21 These are considerably higher than the 73.3% survival rate of persons with non-traumatic SCI in our study. Unfortunately, there is no consensus on the definition of non-traumatic SCI and outcomes may vary on the underlying disease.6, 21
Few other studies have compared survival rates in persons with traumatic and non-traumatic SCI. In one study, non-non-traumatic SCI was a predictor of mortality in univariate analyses, but was not after adjustment for age.22 Another study found no significant differences in risk factors for mortality in non-traumatic and traumatic SCI, other than the effect of age at lesion onset, which was a greater risk factor in the latter group.21
Advanced age at injury is commonly described as risk factor of mortality.1 The strength of the predictive value of age at injury in our study was reduced in the cox regression analyses by the few young persons who died early after SCI, as can be seen in the survival curve (figure 1).
A history of other medical conditions was an independent predictor of death in our study. The additional gathered information showed that 74% had relevant
co-morbidities, suggesting this may have been an important factor, which is confirmed by several other studies.1
Altogether, persons with non-traumatic SCI are generally older and have more co-morbidities, which makes them even more at risk for premature death. The high mortality ratio in persons with SCI may be partly ascribed to the result of a sedentary life style, causing the cardiovascular system to deteriorate more rapidly than in the general population. A healthy life style thus seems extra important for persons with SCI.23
Social support was related to death in bivariate analyses, probably due to the restricted social support that older people experience. In contrast to several other studies,2,9,10 our study did not support the hypothesis that other health, life style and psychosocial factors on admission contributed to death. However, persons with psychiatric diagnosis were excluded, which may have caused a selection bias. Lesion characteristics were also not related to death in our study. Some studies found the same result,3,12,21,22 but other studies did find associations with lesion characteristics.2,5,7,8
The predominant causes of death we found, pulmonary and cardiovascular disease, are a confirmation of the study by Hartkopp et al. and other studies that describe long-term mortality in persons with SCI.2,4,5,22 No death due to suicide occurred in our study, whereas in some other studies it appears to be a common cause of death.2-5 This may be due to the exclusion of persons with psychiatric problems in our study. Furthermore, our sample did not include persons with mild lesions, a group that has been reported to have higher risk on suicide.4 One death occurred due to euthanasia. Euthanasia is illegal in the Netherlands. However, it goes without prosecution since 2002 as long as legal requirements are met. We also found no deaths due to urinary tract complications. Several studies have described a reduction of deaths due to this cause.1,3,8 The improved care and increased awareness of possible urinary tract complications have probably contributed to the absence of urinary tract related deaths in our study.
Pulmonary embolism is described as an early cause of death in older studies.24,25 In our study, strikingly, within the first year after the onset of SCI three young people died of a pulmonary embolism. Two of them used oral anti-coagulation during their admission, indicating they had a higher risk of developing a thromboembolic event. As one person used oral anti-coagulation until death, the dose of oral anti-coagulation may not have been correctly adjusted. Nevertheless, this finding implicates the importance of awareness for deep venous thrombosis and pulmonary embolism.
limitations
Several limitations should be addressed. In-hospital deaths were not included. The study group was relatively small and concerned a selected cohort of persons with trau-matic and non-trautrau-matic SCI. They were expected to remain wheelchair dependent, meaning that persons with mild lesions were excluded. The age criteria for inclusion
probably render an underestimation of mortality when compared with SCI-population based studies. Effects of psychosocial factors on mortality are possibly underestimated as persons with psychiatric problems were excluded.
Non-traumatic SCI may be due to different etiologic mechanisms, which makes mortality data in this group more difficult to compare. No firm conclusions can be drawn from our variable “history of other medical conditions” which included several factors.
Implications
Persons with non-traumatic SCI, a higher age at injury and a history of other medical conditions have a higher risk of dying in the first years after injury. Discussing the importance of a healthy life style with patients in this group should start early and be a routine part of regular follow up after discharge of rehabilitation.
Further research could include a prolonged follow up and a description of in-hospital deaths of persons with SCI in the Netherlands. More research is required on the factors that affect mortality in different non-traumatic etiologies. Awareness of deep venous thrombosis and pulmonary embolism is important to reduce SCI related deaths.
REfERENCES
1. Van den Berg ME, Castellote JM, de Pedro-Cuesta J, Mahillo-Fernandez I. Survival after spinal cord injury: A systematic review. J Neurotrauma 2010;27:1517-1528.
2. Lidal IB, Snekkevik H, Aamodt G, Hjeltnes N, Biering-Sorensen F, Stanghelle JK. Mortality after spinal cord injury in Norway. J Rehabil Med 2007;39:145-151.
3. Soden RJ, Walsh J, Middleton JW, Craven ML, Rutkowski SB, Yeo JD. Causes of death after spinal cord injury. Spinal Cord 2000;38:604-610.
4. Hartkopp A, Bronnum-Hansen H, Seidenschnur AM, Biering-Sorensen F. Survival and cause of death after traumatic spinal cord injury. A long-term epidemiological survey from Denmark.
Spinal Cord 1997;35:76-85.
5. Frankel HL, Coll JR, Charlifue SW, Whiteneck GG, Gardner BP, Jamous MA, et al. Long-term survival in spinal cord injury: A fifty year investigation. Spinal Cord 1998;36:266-274. 6. DeVivo MJ. Epidemiology of traumatic spinal cord injury: Trends and future implications. Spinal
Cord 2012;50:365-372.
7. Strauss DJ, DeVivo MJ, Paculdo DR, Brooks J, Paculdo D. Trends in life expectancy after spinal cord injury. Arch Phys Med Rehabil 2006;87:1079-1085.
8. O’Connor PJ. Survival after spinal cord injury in Australia. Arch Phys Med Rehabil 2005;86:37-47.
9. Krause JS, Zhai Y, Saunders LL, Carter RE. Risk of mortality after spinal cord injury: An 8-year prospective study. Arch Phys Med Rehabil 2009;90:1708-1715.
10. Cao Y, Krause JS, Dipiro N. Risk factors for mortality after spinal cord injury in the USA. Spinal
Cord 2013;51:413-418.
11. Van Asbeck FW, Post MW, Pangalila RF. An epidemiological description of spinal cord injuries in the Netherlands in 1994. Spinal Cord 2000;38:420–424.
12. Samsa GP, Patrick CH, Feussner JR. Long-term survival of veterans with traumatic spinal cord injury. Arch Neurol 1993;50:909-914.
13. De Groot S, Dallmeijer AJ, Post MW, van Asbeck FW, Nene AV, Angenot EL, et al. Demographics of the Dutch multicenter prospective cohort study ‘restoration of mobility in spinal cord injury rehabilitation’. Spinal Cord 2006;44:668-675.
14. Maynard Jr FM, Bracken MB, Creasey G, Ditunno Jr JF, Donovan WH, Ducker TB. International standards for neurological and functional classification of spinal cord injury. American Spinal Injury Association. Spinal Cord 1997;35:266–274.
15. Hall KM, Cohen CE, Wright J, Call M, Werner P. Characteristics of the Functional Independence Measure in traumatic spinal cord injury. Arch Phys Med Rehabil 1999;80:1471-1476
16. Kempen GI, van Eijk LM. The psychometric properties of the SSL-12-I, a short scale for measuring social support in the elderly. Soc Ind Res 1995;35:303–312
17. Forchheimer M, McAweeney M, Tate DG. Use of the SF-36 among persons with spinal cord injury. Am J Phys Med Rehabil 2004;83:390-395
18. Post MW, van Leeuwen CM, van Koppenhagen CF, de Groot S. Validity of the life satisfaction questions, the life satisfaction questionnaire, and the satisfaction with life scale in persons with spinal cord injury. Arch Phys Med Rehabil 2012;93:1832-1837.
19. Osterthun R, Post MW, van Asbeck FW, Dutch-Flemish Spinal Cord Society. Characteristics, length of stay and functional outcome of patients with spinal cord injury in Dutch and Flemish rehabilitation centres. Spinal Cord 2009;47:339-344.
20. Mesard L, Carmody A, Mannarino E, Ruge D. Survival after spinal cord trauma. A life table analysis. Arch Neurol 1978;35:78-83.
21. Ronen J, Goldin D, Bluvshtein V, Fishel B, Gelernter I, Catz A. Survival after nontraumatic spinal cord lesions in Israel. Arch Phys Med Rehabil 2004;85:1499-1502.
22. Garshick E, Kelley A, Cohen SA, Garrison A, Gagnon D, Brown R. A prospective assessment of mortality in chronic spinal cord injury. Spinal Cord 2005;43:408-416.
23. Van der Woude LH, de Groot S, Postema K, Bussmann JB, Janssen TW, ALLRISC, et al. Active LifestyLe rehabilitation interventions in aging spinal cord injury (ALLRISC): A multicentre research program. Disabil Rehabil 2013;35:1097-1103.
24. DeVivo MJ, Black KJ, Stover SL. Causes of death during the first 12 years after spinal cord injury.
Arch Phys Med Rehabil 1993;74:248-254.
25. DeVivo MJ, Kartus PL, Stover SL, Rutt RD, Fine PR. Cause of death for patients with spinal cord injuries. Arch Intern Med 1989;149:1761-1766.
